For Agents
Generate rich link previews (title, description, image, canonical URL) for any URL so agents can render social-style cards in chat, feeds, or bookmarking flows.
Use for: I need to generate a link preview for a URL pasted in chat, I want to fetch Open Graph metadata for an article URL, Get the title and image from a webpage, Retrieve the canonical URL for a shared link
Not supported: Does not handle full article extraction, screenshots, or PDF rendering - use for Open Graph link preview metadata only.
LinkPreview extracts Open Graph metadata and renders link previews from any URL. Given a target URL, the API returns the page title, description, canonical URL, and a representative image, with optional fields for image dimensions, favicon, locale, and content-safety detection on higher-tier plans. Useful for chat apps, social feeds, bookmarking tools, and any product that needs to render rich previews from arbitrary user-submitted links.
Install Jentic One Beta
Jentic One is a self-hosted execution layer for AI agents. It lets your agent call the LinkPreview API, or any other public or private API you need. You set the rules, the agent never sees your credentials, and every call is logged.
Two steps, two machines. Install the instance in a safe environment, then register your agent from wherever it runs.
Step 1: Jentic One Host machine
# On the machine that will host your Jentic One instance:
curl -fsSL "https://jentic.com/install.sh?src=apis&api=%2Fapis%2Flinkpreview.net%2Flinkpreview" | shStep 2: Agent machine
# On the machine where your agent runs (keep this separate from the instance):
curl -fsSL "https://jentic.com/install.sh?src=apis&api=%2Fapis%2Flinkpreview.net%2Flinkpreview" | sh
jentic register # connects your agent to your Jentic One instanceJentic One is in public beta. The setup above keeps your agent separate from the instance, which is what you want before using real credentials: an agent running as the same OS user as Jentic One can read its stored keys directly. Just evaluating? A single local install is fine to start. See the secure deployment guide for the tiers.
What an agent can do with LinkPreview API.
Extract Open Graph title, description, image, and canonical URL from any public URL via GET / or POST /
Return image dimensions and favicon for richer card rendering on higher-tier plans
Detect content safety flags on URLs to gate user-submitted links before rendering
Surface page locale to localise preview cards for multilingual feeds
Authenticate every request with an X-Linkpreview-API-Key header scoped to the account
Patterns agents use LinkPreview API for, with concrete tasks.
★ Chat Link Unfurling
Render rich preview cards when users paste URLs into a chat or messaging app. The LinkPreview API returns title, description, and a representative image in a single GET / call, letting the chat client display a Slack- or Discord-style unfurl. Most calls return in well under a second for cached URLs, making it suitable for in-message rendering.
Call GET /?q=https://example.com/article and return the title, description, and image fields formatted as a chat card.
Bookmark and Reading-List Enrichment
Enrich saved bookmarks and reading-list items with title, description, and a thumbnail so users can scan their saved content visually. POST / accepts a JSON body with the target URL and returns the same metadata, which can be persisted alongside the bookmark record. Suitable for products that ingest dozens to thousands of bookmarks per user.
POST / with body {q: 'https://news.example.com/post-1'} and persist the returned title, description, and image fields with the bookmark.
Social Feed Card Rendering
Render link cards in user-generated social feeds without scraping target pages directly. LinkPreview handles Open Graph, Twitter Card, and oEmbed parsing so the feed product gets a normalised payload regardless of how the source page exposes metadata. Reduces feed-rendering integration time from days to minutes.
For each new feed post, call GET /?q={post.url} and store image, title, and description for card rendering.
AI Agent Link Summarisation
Let an AI agent summarise a URL's metadata before deciding whether to fetch the full page or hand the link to a user. Through Jentic, the agent searches for the LinkPreview operation, loads the schema, and gets a compact metadata payload it can include in chat replies. This avoids unnecessary full-page fetches while still grounding the response in the page's own description.
Search Jentic for 'get a link preview', load the schema, execute GET / with q=<user URL>, and use the description in the agent's reply.
2 endpoints — linkpreview extracts open graph metadata and renders link previews from any url.
METHOD
PATH
DESCRIPTION
/
Get URL preview via query parameter
/
Get URL preview via JSON body
/
Get URL preview via query parameter
/
Get URL preview via JSON body
What agents get from Jentic-routed access to this vendor.
Setup
Wiring LinkPreview by hand means setting up its X-Linkpreview-API-key header auth, attaching it to every request against api.linkpreview.net, and handling your own error and retry logic. Through Jentic you install once, import the LinkPreview API from the API Directory, store the key once, and your agent calls it.
Permission scoping
LinkPreview takes the target URL in the request rather than as a resource id in the URL path, so limit the agent to the operations it needs, such as fetching preview metadata with GET / or POST /. You choose the operations it may call, so the agent is confined to link-preview reads and nothing more.
Credential isolation
Your LinkPreview key is stored once, encrypted, by your own Jentic One instance and injected as the X-Linkpreview-API-key header at execution time. It never enters the agent's prompt, logs, or context.
Intent-based discovery
Agents search Jentic by intent such as 'get a link preview for a URL', and Jentic returns the matching LinkPreview operation with its input schema so the agent calls the right endpoint without browsing the reference docs.
Alternatives and complements available in the Jentic catalogue.
Specific to using LinkPreview API through Jentic.
What authentication does the LinkPreview API use?
LinkPreview uses API key authentication via the X-Linkpreview-API-Key header. Through Jentic, the key is held in your Jentic One instance and injected at request time so agents never receive the raw key in their context.
Can I extract images and titles from any URL with LinkPreview?
Yes. GET / returns title, description, image, and canonical URL for any publicly accessible URL by parsing Open Graph, Twitter Card, and oEmbed tags. POST / accepts the same query in a JSON body when you prefer not to put the URL in a query string.
What are the rate limits for the LinkPreview API?
Rate limits are tiered by plan and are enforced server-side. Free plans are throttled tightly (typically a few requests per minute) while paid plans expand to thousands per day. Build agents to cache previews per URL rather than re-fetching on every render.
How do I generate a link preview through Jentic?
Run pip install jentic, then search for 'get a link preview' to find GET / on linkpreview.net. Load the schema, supply q=<target URL>, and execute. The returned metadata can be passed straight into a chat card or feed renderer.
Does LinkPreview support content safety detection?
Higher-tier plans return a content-safety flag alongside the standard metadata, which is useful for moderating user-submitted links before rendering. Check the response for the safety field on each preview and gate rendering accordingly.
Can I limit what my agent is allowed to do with the LinkPreview API?
Yes. Because Jentic One is self-hosted by you, your own rules decide which LinkPreview operations and credentials the agent may use. The LinkPreview API takes the target URL in the request rather than as a resource id in the path, and it exposes only two read operations, GET / and POST /, so you can allow just the preview-metadata calls your agent needs. That confines the agent to fetching Open Graph title, description, image, and canonical URL, and nothing more.
GET STARTED